End of training

README.md

@@ -2,9 +2,15 @@
 license: llama2
 base_model: llama-2-nl/Llama-2-7b-hf-lora-tokentrans-sft
 tags:
+- alignment-handbook
 - trl
 - dpo
 - generated_from_trainer
+- trl
+- dpo
+- generated_from_trainer
+datasets:
+- BramVanroy/ultra_feedback_dutch
 model-index:
 - name: Llama-2-7b-hf-lora-tokentrans-it
   results: []
@@ -15,17 +21,17 @@ should probably proofread and complete it, then remove this comment. -->
 
 # Llama-2-7b-hf-lora-tokentrans-it
 
-This model is a fine-tuned version of [llama-2-nl/Llama-2-7b-hf-lora-tokentrans-sft](https://huggingface.co/llama-2-nl/Llama-2-7b-hf-lora-tokentrans-sft) on an unknown dataset.
+This model is a fine-tuned version of [llama-2-nl/Llama-2-7b-hf-lora-tokentrans-sft](https://huggingface.co/llama-2-nl/Llama-2-7b-hf-lora-tokentrans-sft) on the BramVanroy/ultra_feedback_dutch dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.3909
-- Rewards/chosen: 0.1789
-- Rewards/rejected: -0.6788
-- Rewards/accuracies: 0.9396
-- Rewards/margins: 0.8577
-- Logps/rejected: -556.9485
-- Logps/chosen: -600.6838
-- Logits/rejected: 1.1685
-- Logits/chosen: 1.5742
+- Loss: 0.3913
+- Rewards/chosen: 0.1776
+- Rewards/rejected: -0.6740
+- Rewards/accuracies: 0.9418
+- Rewards/margins: 0.8516
+- Logps/rejected: -556.9005
+- Logps/chosen: -600.6971
+- Logits/rejected: 1.1696
+- Logits/chosen: 1.5756
 
 ## Model description
 

all_results.json

@@ -1,5 +1,18 @@
 {
     "epoch": 0.9990049751243781,
+    "eval_logits/chosen": 1.5755633115768433,
+    "eval_logits/rejected": 1.1696137189865112,
+    "eval_logps/chosen": -600.6971435546875,
+    "eval_logps/rejected": -556.9004516601562,
+    "eval_loss": 0.3912544250488281,
+    "eval_rewards/accuracies": 0.9417910575866699,
+    "eval_rewards/chosen": 0.17757247388362885,
+    "eval_rewards/margins": 0.851599395275116,
+    "eval_rewards/rejected": -0.6740269660949707,
+    "eval_runtime": 879.0322,
+    "eval_samples": 5359,
+    "eval_samples_per_second": 6.096,
+    "eval_steps_per_second": 0.381,
     "total_flos": 0.0,
     "train_loss": 0.45811287039621257,
     "train_runtime": 34472.405,

config.json

@@ -0,0 +1,36 @@
+{
+  "_name_or_path": "llama-2-nl/Llama-2-7b-hf-lora-tokentrans-sft",
+  "additional_special_tokens_ids": [],
+  "architectures": [
+    "LlamaForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 0,
+  "cls_token_id": 0,
+  "eos_token_id": 3,
+  "hidden_act": "silu",
+  "hidden_size": 4096,
+  "initializer_range": 0.02,
+  "intermediate_size": 11008,
+  "mask_token_id": 4,
+  "max_position_embeddings": 4096,
+  "mlp_bias": false,
+  "model_type": "llama",
+  "num_attention_heads": 32,
+  "num_hidden_layers": 32,
+  "num_key_value_heads": 32,
+  "pad_token_id": 1,
+  "pretraining_tp": 1,
+  "rms_norm_eps": 1e-05,
+  "rope_scaling": null,
+  "rope_theta": 10000.0,
+  "sep_token_id": 3,
+  "tie_word_embeddings": false,
+  "tokenizer_class": "RobertaTokenizerFast",
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.40.1",
+  "unk_token_id": 2,
+  "use_cache": true,
+  "vocab_size": 50000
+}

eval_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 0.9990049751243781,
+    "eval_logits/chosen": 1.5755633115768433,
+    "eval_logits/rejected": 1.1696137189865112,
+    "eval_logps/chosen": -600.6971435546875,
+    "eval_logps/rejected": -556.9004516601562,
+    "eval_loss": 0.3912544250488281,
+    "eval_rewards/accuracies": 0.9417910575866699,
+    "eval_rewards/chosen": 0.17757247388362885,
+    "eval_rewards/margins": 0.851599395275116,
+    "eval_rewards/rejected": -0.6740269660949707,
+    "eval_runtime": 879.0322,
+    "eval_samples": 5359,
+    "eval_samples_per_second": 6.096,
+    "eval_steps_per_second": 0.381
+}

runs/May21_11-10-11_gpu538.dodrio.os/events.out.tfevents.1716318836.gpu538.dodrio.os.3651941.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e4a7f0937d07155c4f6af84806e1d62f804ef3373a87c699c8ae963f44aefadd
+size 828